Uniform fuel arrays and uniform fires do not occur. All fire behavior in 

 woody fuel arrays is nonuniform. Fire behavior becomes uniform by definition. As 

 a point of comparison, most people consider fire spreading over a sheet of paper 

 as uniform. But, it may be viewed by some as nonuniform by focusing on the elemen- 

 tal combustion process. Certainly, fuel uniformity does not exist in natural woody 

 fuel arrays. The arrays are comprised of separated particles. Pons (1946) described 

 fire spread as a series of ignitions. Fire spread appears to follow this description 

 on the scale of the fuel separation. The flame ignites the particle and then spreads 

 along that particle until it reaches another particle. The flame may ignite the next 

 particle by flame contact or the particle may ignite after radiation from the flame 

 front has preheated the fuel sufficiently to drive off combustible gases that bum 

 when in contact with flame. 



A fire appears uniform when the particle separation is small compared to the 

 flame size (or the range of influence of the flame). Individual flames coalesce 

 into a solid moving front and the fire spreads in response to the bulk fuel proper- 

 ties of the array--load, particle size distribution, and depth. These properties 

 are defined within a volume element, averaging out the small variations of particle 

 separation. Averaging is not limiting so long as the volume element encountered by 

 the flame is small compared to the size or range of influence of the flame. It is 

 this level of nonuniformity that we address. tVind and slope can act to increase the 

 size of the flame and orient it so that the spatial variation of bulk fuel proper- 

 ties is small compared to the area affected by the flame. Thus, a nonuniform fire 

 can become uniform in the presence of wind or slope. 



Although resolution is limited by the cell size, the introduction of wind 

 alters the overall perspective of resolution. Wind increases the size of the area 

 sustaining active combustion and therefore increases the error of locating the 

 fire. Consequently, in the presence of wind, the size of the cell can be increased 

 without altering accuracy in locating the fire. 



The model is applied to two examples of nonuniformity: slash, residue left 

 after tree harvesting, and a mixed community of grass and sagebrush. (Slash is 

 more uniform than a mixed community of grass and brush.) The reader should expect 

 to gain an appreciation for the problem of defining fire nonuniformity, developing 

 a model of fire behavior that responds to fuel nonuniformity, and an appreciation 

 of the kind and form of the results obtained from the nonuniformity model. Attempts 

 are not made to validate the model. Consequently, the rigors of replication are 

 replaced with a logical flow of fire behavior concepts --concepts derived from the 

 uniform fire behavior model. The initial advantage is a consistent manner of 

 handling field data that is applied to nonuniform fire behavior. 



The response of a spreading fire to the bulk properties of woody fuel arrays-- 

 as found in forest fuels--has been investigated by Fons (1946) , Thomas and Simms 

 (1963), Anderson (1968, 1969), Frandsen (1971), Steward (1971), and Rothermel (1972). 

 Rothermel incorporated fuel parameters (load, size, depth) and the fuel interactions 

 into a model of fire spread through a continuous fuel array. Although the fuel 

 array is continuous, it may be heterogeneous in size and type. Live and dead fuel 

 may be included if mixed in the same stratum. 



Fuel parameters for the uniform fire spread model are categorized as living or 

 dead, and averaged within a specific set of fuel size classes. The fuel array is 

 assumed to be continuous. As a consequence, the model given by Rothermel (1972) 

 predicts well for spatially continuous fuels and becomes increasingly less accurate 

 as fuel discontinuity increases. To properly assess fire behavior, continuity must 

 be included as an essential parameter in the mechanism of fire spread. Brown (1966) 

 described the problem of continuity as follows: 



2 



